Search

in press Accurate classification of melanocytic tumors is important for prognostic evaluation, treatment and follow-up protocols of patients. The majority of melanocytic proliferations can be classified solely based on clinical and pathological criteria, however in select cases a definitive diagnostic assessment remains challenging and additional diagnostic biomarkers would be advantageous. We analyzed melanomas, nevi, Spitz nevi and atypical spitzoid tumors using parallel sequencing (exons of 611 genes and 507 gene translocation analysis) and methylation arrays (850k Illumina EPIC). By combining detailed genetic and epigenetic analysis with reference-based and reference-free DNA methylome deconvolution we compared Spitz nevi to nevi and melanoma and assessed the potential for these methods in classifying challenging spitzoid tumors. Results were correlated with clinical and histologic features. Spitz nevi were found to cluster independently of nevi and melanoma and demonstrated a different mutation profile. Multiple copy number alterations and TERT promoter mutations were identified only in melanomas. Genome-wide methylation in Spitz nevi was comparable to benign nevi while the Leukocytes UnMethylation for Purity (LUMP) algorithm in Spitz nevi was comparable to melanoma. Histologically difficult to classify Spitz tumor cases were assessed which, based on methylation arrays, clustered between Spitz nevi and melanoma and in terms of genetic profile or copy number variations demonstrated worrisome features suggesting a malignant neoplasm. Comprehensive sequencing and methylation analysis verify Spitz nevi as an independent melanocytic entity distinct from both nevi and melanoma. Combined genetic and methylation assays can offer additional insights in diagnosing difficult to classify Spitzoid tumors.

Purpose: Recent studies have demonstrated HLA class II (HLA-II)-dependent killing of melanoma cells by cytotoxic CD4 T cells. We investigated evolution of HLA-II-loss tumors that escape cytotoxic CD4 T cell activity and contribute to immunotherapy resistance. Experimental Design: Melanoma cells from longitudinal metastases were studied for constitutive and interferon-inducible HLA-II expression, sensitivity towards autologous CD4 T cells, and immune evasion by HLA-II loss. Clinical significance of HLA-II-low tumors was determined by analysis of transcriptomic data sets from patients with immune checkpoint blockade (ICB). Results: Analysis of longitudinal samples revealed strong inter-metastatic heterogeneity in melanoma cell-intrinsic HLA-II expression and subclonal HLA-II loss. Tumor cells from early lesions either constitutively expressed HLA-II, sensitizing to cytotoxic CD4 T cells, or induced HLA-II and gained CD4 T cell sensitivity in the presence of IFNγ. In contrast, late outgrowing subclones displayed a stable CD4 T cell-resistant HLA-II-loss phenotype. These cells lacked not only constitutive but also IFNγ-inducible HLA-II due to JAK1/2-STAT1 pathway inactivation. Coevolution of JAK1/2 deficiency and HLA-II loss established melanoma cross-resistance to IFNγ and CD4 T cells, as detected in distinct stage IV metastases. In line with their immune-evasive phenotype, HLA-II-low melanomas showed reduced CD4 T cell infiltrates and correlated with disease progression under ICB. Conclusions: Our study links melanoma resistance to CD4 T cells, IFNγ, and ICB at the level of HLA-II, highlighting the significance of tumor cell-intrinsic HLA-II antigen presentation in disease control and calling for strategies to overcome its downregulation for improvement of patient outcome

The success of cancer immunotherapy is limited by resistance to immune checkpoint blockade. We therefore conducted a genetic screen to identify genes that mediated resistance against CTLs in anti–PD-L1 treatment–refractory human tumors. Using PD-L1–positive multiple myeloma cells cocultured with tumor-reactive bone marrow–infiltrating CTL as a model, we identified calcium/calmodulin-dependent protein kinase 1D (CAMK1D) as a key modulator of tumor-intrinsic immune resistance. CAMK1D was coexpressed with PD-L1 in anti–PD-L1/PD-1 treatment–refractory cancer types and correlated with poor prognosis in these tumors. CAMK1D was activated by CTL through Fas-receptor stimulation, which led to CAMK1D binding to and phosphorylating caspase-3, -6, and -7, inhibiting their activation and function. Consistently, CAMK1D mediated immune resistance of murine colorectal cancer cells in vivo. The pharmacologic inhibition of CAMK1D, on the other hand, restored the sensitivity toward Fas-ligand treatment in multiple myeloma and uveal melanoma cells in vitro. Thus, rapid inhibition of the terminal apoptotic cascade by CAMK1D expressed in anti–PD-L1–refractory tumors via T-cell recognition may have contributed to tumor immune resistance.

siRNA library

Supplementary Table 2

Supplementary Figures 1-7

Tab. S1: Genes covered in the applied sequencing Panel; Tab. S2: List of vaccine peptides

Purpose: CD8+ T lymphocytes can kill autologous melanoma cells, but their activity is impaired when poorly immunogenic tumor phenotypes evolve in the course of disease progression. Here, we analyzed three consecutive melanoma lesions obtained within one year of developing stage IV disease for their recognition by autologous T cells.Experimental Design: One skin (Ma-Mel-48a) and two lymph node (Ma-Mel-48b, Ma-Mel-48c) metastases were analyzed for T-cell infiltration. Melanoma cell lines established from the respective lesions were characterized, determining the T-cell–stimulatory capacity, expression of surface molecules involved in T-cell activation, and specific genetic alterations affecting the tumor–T-cell interaction.Results: Metastases Ma-Mel-48a and Ma-Mel-48b, in contrast with Ma-Mel-48c, were infiltrated by T cells. The T-cell–stimulatory capacity was found to be strong for Ma-Mel-48a, lower for Ma-Mel-48b, and completely abrogated for Ma-Mel-48c cells. The latter proved to be HLA class I–negative due to an inactivating mutation in one allele of the beta-2-microglobulin (B2M) gene and concomitant loss of the other allele by a deletion on chromosome 15q. The same deletion was already present in Ma-Mel-48a and Ma-Mel-48b cells, pointing to an early acquired genetic event predisposing to development of β2m deficiency. Notably, the same chronology of genetic alterations was also observed in a second β2m-deficient melanoma model.Conclusion: Our study reveals a progressive loss in melanoma immunogenicity during the course of metastatic disease. The genetic evolvement of T-cell resistance suggests screening tumors for genetic alterations affecting immunogenicity could be clinically relevant in terms of predicting patient responses to T-cell–based immunotherapy. Clin Cancer Res; 20(24); 6593–604. ©2014 AACR.

Supplementary Fig. S4. Real-time proliferation of Ma-Mel-48 cell lines was determined in an xCelligence device. 1,5 x 104 cells were seeded per well and proliferation was determined over time. Representative data from one of three independent experiments are depicted

Melanoma often recurs after a latency period of several years, presenting a T cell–edited phenotype that reflects a role for CD8+ T cells in maintaining metastatic latency. Here, we report an investigation of a patient with multiple recurrent lesions, where poorly immunogenic melanoma phenotypes were found to evolve in the presence of autologous tumor antigen–specific CD8+ T cells. Melanoma cells from two of three late recurrent metastases, developing within a 6-year latency period, lacked HLA class I expression. CD8+ T cell–resistant, HLA class I–negative tumor cells became clinically apparent 1.5 and 6 years into stage IV disease. Genome profiling by SNP arrays revealed that HLA class I loss in both metastases originated from a shared chromosome 15q alteration and independently acquired focal B2M gene deletions. A third HLA class I haplotype-deficient lesion developed in year 3 of stage IV disease that acquired resistance toward dominant CD8+ T-cell clonotypes targeting stage III tumor cells. At an early stage, melanoma cells showed a dedifferentiated c-Junhigh/MITFlow phenotype, possibly associated with immunosuppression, which contrasted with a c-Junlow/MITFhigh phenotype of T cell–edited tumor cells derived from late metastases. In summary, our work shows how tumor recurrences after long-term latency evolve toward T-cell resistance by independent genetic events, as a means for immune escape and immunotherapeutic resistance. Cancer Res; 76(15); 4347–58. ©2016 AACR.

Additional information about experimental techniques and materials relevant to this study

Legends

Supplementary Figures S1-S6. Supplementary Fig. S1, localization of T cells in the periphery of metastasis Ma-Mel-86f. Supplementary Fig. S2, re-expression of HLA class I on B2M-transfected melanoma cells. Supplementary Fig. S3, no enrichment of CD4+ T cells and NK cells in metastasis Ma-Mel-86f. Supplementary Fig. S4, similar PD-L1 expression on Ma-Mel-86a and Ma-Mel-86c cells. Supplementary Fig. S5, autologous CD8+ T cells kill Ma-Mel-86c cells. Supplementary Fig. S6, autologous CD8+ T cells kill Ma-Mel-86a cells.

Background Recent studies have successfully demonstrated the use of deep-learning algorithms for dermatologist-level classification of suspicious lesions by the use of excessive proprietary image databases and limited numbers of dermatologists. For the first time, the performance of a deep-learning algorithm trained by open-source images exclusively is compared to a large number of dermatologists covering all levels within the clinical hierarchy. Methods We used methods from enhanced deep learning to train a convolutional neural network (CNN) with 12,378 open-source dermoscopic images. We used 100 images to compare the performance of the CNN to that of the 157 dermatologists from 12 university hospitals in Germany. Outperformance of dermatologists by the deep neural network was measured in terms of sensitivity, specificity and receiver operating characteristics. Findings The mean sensitivity and specificity achieved by the dermatologists with dermoscopic images was 74.1% (range 40.0%–100%) and 60% (range 21.3%–91.3%), respectively. At a mean sensitivity of 74.1%, the CNN exhibited a mean specificity of 86.5% (range 70.8%–91.3%). At a mean specificity of 60%, a mean sensitivity of 87.5% (range 80%–95%) was achieved by our algorithm. Among the dermatologists, the chief physicians showed the highest mean specificity of 69.2% at a mean sensitivity of 73.3%. With the same high specificity of 69.2%, the CNN had a mean sensitivity of 84.5%. Interpretation A CNN trained by open-source images exclusively outperformed 136 of the 157 dermatologists and all the different levels of experience (from junior to chief physicians) in terms of average specificity and sensitivity.

Background: Recent studies have demonstrated the use of convolutional neural networks (CNNs) to classify images of melanoma with accuracies comparable to those achieved by board-certified dermatologists. However, the performance of a CNN exclusively trained with dermoscopic images in a clinical image classification task in direct competition with a large number of dermatologists has not been measured to date. This study compares the performance of a convolutional neuronal network trained with dermoscopic images exclusively for identifying melanoma in clinical photographs with the manual grading of the same images by dermatologists. Methods: We compared automatic digital melanoma classification with the performance of 145 dermatologists of 12 German university hospitals. We used methods from enhanced deep learning to train a CNN with 12,378 open-source dermoscopic images. We used 100 clinical images to compare the performance of the CNN to that of the dermatologists. Dermatologists were compared with the deep neural network in terms of sensitivity, specificity and receiver operating characteristics. Findings: The mean sensitivity and specificity achieved by the dermatologists with clinical images was 89.4% (range: 55.0%-100%) and 64.4% (range: 22.5%-92.5%). At the same sensitivity, the CNN exhibited a mean specificity of 68.2% (range 47.5%-86.25%). Among the dermatologists, the attendings showed the highest mean sensitivity of 92.8% at a mean specificity of 57.7%. With the same high sensitivity of 92.8%, the CNN had a mean specificity of 61.1%. Interpretation: For the first time, dermatologist-level image classification was achieved on a clinical image classification task without training on clinical images. The CNN had a smaller variance of results indicating a higher robustness of computer vision compared with human assessment for dermatologic image classification tasks. (C) 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Myeloid derived suppressor cells (MDSCs) suppress innate and adaptive immunity, thereby limiting anti-tumor immune responses in cancer patients. In patients with advanced melanoma, the phenotype and function of MDSCs remains controversial. In our study, we further explored two distinct subpopulations of MDSCs and investigated the impact of Vemurafenib on these cells. Flow cytometry analysis revealed that in comparison to healthy donors and patients with localized disease, PBMCs from patients with metastatic melanoma showed an increased frequency of CD14(+) HLA-DR(-/low) monocytic MDSCs (moMDSCs) and of a previously unrecognized population of CD14(-) CD66b(+) Arginase1(+) granulocytic MDSCs (grMDSCs). In vitro, both populations suppressed autologous T-cell proliferation, which was tested in CFSE-based proliferation assays. Vemurafenib treatment of melanoma patients reduced the frequency of both moMDSCs and grMDSCs. According to our in vivo finding, conditioned medium (CM) from Vemurafenib treated melanoma cells was less active in inducing moMDSCs in vitro than CM from untreated melanoma cells. In conclusion, patients with advanced melanoma show increased levels of moMDSCs, and of a population of CD14(-) CD66b(+) Arginase1(+) grMDSCs. Both MDSCs are distinct populations capable of suppressing autologous T-cell responses independently of each other. In vitro as well as in vivo, Vemurafenib inhibits the generation of human moMDSCs. Thus, Vemurafenib decreases immunosuppression in patients with advanced melanoma, indicating its potential as part of future immunotherapies.

Dendritic cells (DC) play an important role as antigen-presenting cells in T cell stimulation. Interestingly, a number of recent studies also imply DC as critical accessory cells in B cell activation, isotype switching and plasma blast maintenance. Here we use the conditional in vivo ablation of CD11c(high) DC to investigate the role of these cells in T-independent type 2 immune responses. We show that CD11c(high) DC are dispensable for the initiation and maintenance of a primary immune response against the T-independent type 2 antigen (4-hydroxy-3-nirophenyl)acetyl-Ficoll. Our results suggest that support for plasma cell formation in T cell-independent immune responses can be provided by non-DC such as stromal cells, or is independent of external signals. Interestingly, we found plasma blasts to express CD11c and to be diphtheria toxin-sensitive in CD11c-diphtheria toxin receptor-transgenic mice, providing a unique tool for future analysis of in vivo aspects of plasma cell biology.